Hint 1: Backing up the data folder of a running MySQL instance is not a good idea. You want to either stop that service prior to backing up or, even better, make a database dump and back that up at the same time you backup the rest of the application files to make sure you have a consistant set of files that can be successfully restored. Which you regularily want to test anyway.
Hint 2: Consider using a single restic repo if you expect to be backing up the same data multiple times. That way you can enjoy the benefits of restic’s deduplication functionality and save a lot of space on the backup locations.
That will work. Use one repo if deduplication is possible and wanted or multiple repos if you want to be able to, say, one day split repos to different drives. It depends on your use case.
@adamol try a restore and see if you can easily identify the different VM.
If not then I would suggest to look at tagging and hostname during backup. These are the major differentiators for restic to filter snapshots.
UPDATE: I forgot there are also options for --path at snapshots and restore commands so you’d likely have no problems at all.
@nicnab@Ilya Hello, everyone. I would appreciate your advice. I still don’t understand why you recommend a single repository. What’s the point? He’ll get the same deduplicated storage, but with more flexible settings, right? The number of storage units doesn’t affect the amount of space occupied.
What do you mean by “storage unit”? If you have two restic repos for two hosts with identical data, you need twice the space compared to a single, shared repo. Of course two hosts will rarely be exactly the same but even if they have 10% overlay in their data, you save 10% storage space. The only downside is that it’s harder to separate the two hosts to different storage spaces later. It’s not impossible but you have to do some work.